Artificial Intelligence-enhanced electrocardiogram analysis: the need to develop image-based algorithms

31/01/24

Take Home Messages
  • Contemporary artificial intelligence-enhanced electrocardiogram algorithms exhibit levels of diagnostic accuracy comparable with expert clinicians.
  • The ability of artificial intelligence-enhanced algorithms to detect patterns unrecognisable by the human eye offers true promise to expand the utility of the electrocardiogram.
  • Most algorithms have been developed using digitised electrocardiogram signal data and are therefore unable to interpret paper-based electrocardiograms.
  • Image-based algorithms can facilitate paper-based artificial intelligence-enhanced electrocardiogram analysis ensuring algorithm availability to all healthcare professionals.
Introduction

Artificial intelligence has the potential to revolutionise the care pathway for patients with cardiovascular disease through its ability to rapidly analyse the vast volumes of data acquired during routine clinical care (Figure 1).(1)

The electrocardiogram is a widely available diagnostic tool which records the heart’s electrical activity as a graph of voltage versus  time.(2) Numerous studies have already demonstrated the potential artificial intelligence-enhanced electrocardiography has to improve patient care.(3) The aim of this editorial is to summarise the latest developments in artificial intelligence-enhanced electrocardiography and to discuss potential barriers limiting integration into clinical practice.